面向爆管检测的供水管网水压监测点多目标优化  

Multi-objective Optimization of Water Pressure Monitoring Points for Pipe Burst Detection in Water Supply Networks

作　　者：赵文轩 杜坤[1] 孟繁艺 罗雄武 周明[1] 宋志刚[1] ZHAO Wen-xuan;DU Kun;MENG Fan-yi;LUo Xiong-wu;ZHOU Ming;SONG Zhi-gang(Faculty of Civil Engineering and Mechanics,Kunming University of Science and Technology,Kunming 650500,China;Yunnan Design Institute Group Co.Ltd.,Kunming 650100,China;Kunming Guanfang Architecture&Design Inc.,Kunming 650051,China)

机构地区：[1]昆明理工大学建筑工程学院,云南昆明650500 [2]云南省设计院集团有限公司,云南昆明650100 [3]昆明官房建筑设计有限公司,云南昆明650051

出　　处：《中国给水排水》2023年第7期57-62,共6页China Water & Wastewater

基　　金：国家自然科学基金资助项目(52260011);云南省重点领域科技计划项目(202003AC100001)。

摘　　要：针对单目标优化算法在水压监测点优化布置模型求解过程中计算效率低、解集质量差的问题,提出了一种多目标非支配差分进化算法(NSDE)。首先,对各管段中间添加的节点赋予爆管流量并依次进行模拟,根据各节点压力变化值生成爆管事件判断矩阵;然后,以监测点个数最小化和爆管检测漏损量最大化为目标函数构建监测点优化布置数学模型,利用NSDE算法进行求解;最后对Pareto解集中监测点布局特征展开分析。以net3供水管网为例,采用单目标遗传算法、多目标非支配遗传算法(NSGA-Ⅱ)和NSDE算法分别对模型进行求解,结果表明多目标NSDE算法计算效率更高且优化解质量更好,其Pareto解集对应的多种布局方案可为实际工程中监测点布置提供灵活选择。The single objective optimization algorithm has low computational efficiency and poor solution set quality in the process of solving the optimal arrangement model of water pressure monitoring points.A multi-objective non-dominated sorting differential evolution(NSDE)algorithm was proposed to solve these problems.Firstly,the nodes added in the middle of each pipe segment were distributed a pipe burst flow and the pipe burst simulation was carried out successively,and the pipe burst event judgment matrix was generated according to the pressure change value of each node.Then,a mathematical model for optimizing the arrangement of monitoring points was constructed with the objective function of minimizing the number of monitoring points and maximizing the leakage volume of pipe burst detection,and the solution was obtained by using NSDE algorithm.Finally,the arrangement characteristics of monitoring points in Pareto solution were analyzed and discussed.The model was solved by single objective genetic algorithm,multi-objective non-dominated sorting genetic algorithm(NSGA-Ⅱ)and NSDE algorithm respectively exemplified by the net3 water supply network.The multi-objective NSDE algorithm has higher computational efficiency and better optimal solution set quality,and the multiple layout schemes corresponding to its Pareto solution sets provide flexibility for monitoring site arrangement in actual projects.

关 键 词：供水管网 爆管 水压监测点 多目标优化 非支配差分进化算法 

分 类 号：TU991[建筑科学—市政工程]